For example, in order to detect a drowsy driving in a vehicle, detecting a drowsy state from a driver's heart rate is suggested (for example, see Patent Literature 1 or 2).
In the Pattern Literature 1, an average value of RRI (R-R Interval) value detected by a heartbeat sensor during wakefulness and a predetermined multiple of an integral value of the RRI value exceeding the average value are set as a threshold value. When a value after integrating the RRI value exceeding the average value exceeds the threshold value, it is determined to be dozing.
In the patent Literature 2, a heartbeat interval data in which the RRI is arranged time-serially based on a heartbeat waveform of a subject in a rest state measured by a sensor is obtained. Then, frequency analysis of the heartbeat interval data is performed, and the frequency analysis result of heart fluctuation including power spectral density (PSD) in a frequency at a certain time and total power (TP) of autonomic nerve is obtained. Next, the frequency analysis result is set as an initial state, a sleepiness scale in which an origin of sleepiness position and an origin of a waking position are set is determined based on the estimated sleepiness position and the estimated waking position including the initial state. Then, a sleepiness of a subject is determined based on the sleepiness scale.